Particles and their behaviour · Investigation
Testing the model: does it explain everything?
You have spent five lessons being shown how well the particle model works. Here are seven things it has to account for. Three of them it cannot. What should be done about that?
Start here
Ice floats.
The model says a solid is particles packed tightly in rows, and a liquid is the same particles jumbled and slightly further apart. Follow that through and every solid should be denser than its own liquid, and should sink in it. Almost all of them do. Water does not — ice floats, which is why lakes freeze from the top and fish survive the winter.
The model has just made a wrong prediction. Commit to what that means.
Hold your answer. By the end of this lesson you will have seen the model succeed four times and fail three, and the right thing to do about that is the actual content of the lesson — not a tidy verdict either way.
Every lesson so far has handed you evidence that fits. That is not how the model was tested and it is not how you should judge it. A model is only worth anything if you know where its edges are — so this lesson goes looking for them deliberately.
The evidence bench · judge each one
Seven observations. Does the model handle it?
0 of 7 judged
Decide before you open the verdict. You are not being marked on the sorting — you are being asked to commit, so that a surprise can land.
Observation 1
A gas can be squashed into a fraction of its volume; a liquid hardly squashes at all.
The model handles this
Handled completely. A gas is mostly empty space and a liquid is not. Nothing else needs to be added to the model to get this right.
Observation 2
A smell spreads through a room. Draughts and convection carry it most of the way; diffusion covers the last stretch.
The model handles this
Handled completely. Random movement plus collisions gives the spreading — and explains why diffusion on its own would be far too slow to do the whole job.
Observation 3
Ice floats on water. Almost every other solid sinks in its own liquid.
The model cannot do this
Fails. The model predicts that a solid — packed and ordered — must be denser than the same substance jumbled and looser. For water it is the other way round. Identical spheres cannot produce this; you need particles with a shape, that hold each other at arm’s length in a fixed pattern.
Observation 4
Seal a melting ice cube in a bag and the mass does not change by a milligram.
The model handles this
Handled completely. Particles are neither made nor destroyed by a change of state — they are only rearranged.
Observation 5
Diamond and graphite are both nothing but carbon. Diamond is the hardest natural substance known; graphite is soft enough to write with.
The model cannot do this
Fails, and badly. In this model the particles are identical, so the same substance can only have one set of properties. To explain this you need the particles to be joined together in different arrangements — which means particles that can bond, and the model has no bonds in it.
Observation 6
A rubber band stretches to five times its length and snaps back. A glass rod of the same thickness shatters.
The model cannot do this
Fails. Loose spheres sliding past each other cannot stretch and recoil. Rubber needs particles joined into long tangled chains that straighten out and spring back — a structure this model cannot represent at all.
Observation 7
A sealed helium balloon is noticeably smaller after three days, even though it has no hole.
The model handles this
Handled completely — and it is a good test, because the answer is not obvious. The rubber is itself made of particles with gaps between them, and helium particles are small enough to work their way through.
The three failures have one thing in common, and it is worth naming: every one of them needs the particles to be different from each other, or to be joined together in some particular way. The model you have is a model of identical featureless spheres. That is exactly the assumption that is about to be replaced.
The verdict · this is the real question
Three failures. What should be done?
Commit to one. Then read what scientists actually did, which is on the record and is not a matter of opinion.
That is the strictest possible standard, and no model in science survives it. Applied consistently it would leave you with nothing to think with.
This is what was actually done, and it is the answer.
Ice floating is not a curiosity — it is why lakes do not freeze solid and why life survives winter. And the exceptions are where the next model came from, so ignoring them costs you the discovery.
A word change fixes nothing. A theory that predicts ice will sink has exactly the same problem as a model that does.
Nobody threw it away. The particle model is still used every day, by everyone, including the people who know exactly where it breaks — because for melting, pressure, diffusion and dissolving it gives the right answer with almost no effort. What happened instead is that the failures were treated as a map: each one marked a place where a better model was needed, and each one eventually got built. Ice floating was explained once particles could be shaped and could pull on each other in particular directions. Diamond and graphite were explained once particles could be joined in patterns. Rubber was explained once particles could be long chains.
A wrong prediction is not a disgrace. It is the most useful thing a model can produce, because it tells you where to look next.
Key fact
A scientific model is judged by what it explains and where it fails — not by being completely true. Knowing a model's limits is part of understanding it.
Not a straight line
The model has already been replaced four times
Each one of these was, in its day, what "everyone knew". Each was overturned by evidence, and each left something behind that is still in use.
The uncuttable
Matter cannot be divided forever — there must be a smallest piece.
Reached by pure argument, with no experiment behind it and no way to test it. For two thousand years it sat alongside the rival view that matter is continuous, and there was nothing to choose between them.
What broke it: Nothing broke it — which was the problem. An idea that cannot be tested cannot win, and it did not, for twenty centuries.
Solid spheres
Every element is made of identical solid atoms that cannot be split, created or destroyed.
The first version with numbers attached: fixed proportions by mass, whole-number ratios, mass conserved through reactions. This is the model you have been using all unit.
What broke it: Thomson found electrons — pieces knocked off an atom that was supposed to have no pieces. And atoms of one element turned out to have different masses.
Plum pudding
An atom is a ball of positive charge with tiny negative electrons dotted through it.
Kept everything Dalton got right about reactions, and added the one thing he could not have known: atoms have parts, and one of those parts carries charge.
What broke it: Rutherford fired alpha particles at gold foil and a few bounced straight back — impossible if the positive charge were spread thinly through the whole atom.
The nucleus
Almost all the mass sits in a tiny dense nucleus, with electrons somewhere around it and empty space in between.
An atom turns out to be overwhelmingly empty. If the nucleus were a marble on the centre spot of a football pitch, the nearest electron would be somewhere in the stands.
What broke it: The maths said the orbiting electrons should spiral into the nucleus within a fraction of a second. Every atom in existence should have collapsed already.
Shells, then clouds
Electrons are restricted to particular energy levels, and are better described as clouds of probability than as balls.
This is the model behind the periodic table, chemical bonding and every reaction you will meet at GCSE and beyond.
What broke it: Nothing yet, for chemistry. It has limits of its own at very high energies, and the search for what lies past them is a live field of research right now.
Think again
“Once scientists agree on something, it is settled — that is what makes it science.”
Commit before you read on.
The opposite is nearer the truth. What makes an idea scientific is that it is the kind of thing evidence could overturn — and the timeline above is five overturnings in a row, each by people who were not being careless. Dalton was not sloppy; he was working with the evidence he had.
This does not mean nothing is reliable. The particle model has survived two centuries of people trying to break it, and the parts that survived are about as solid as human knowledge gets. "Open to revision" and "as good as we have got" are not opposites — they are the same sentence, said twice.
Mastery ladder
Not started yet.
Rungs 3 and 4 you mark yourself.
Rung 1 · Recall
Name one thing the simple particle model cannot explain.
Rung 2 · The one that catches people
The particle model makes a wrong prediction about ice. What is the right response?
Rung 3 · Explain
The particle model cannot explain why diamond and graphite — both pure carbon — have completely different properties. Explain why not, and say what would have to be added to the model to fix it.
Rung 4 · Take it somewhere new
A student reads the timeline and concludes: "Scientists keep being wrong, so there is no point trusting what they say now." Reply to them, using at least one specific example from this lesson.
Key note
The particle model explains melting, pressure, diffusion and dissolving, and fails on ice floating, diamond versus graphite, and stretchy materials. All three failures come from the same assumption — that particles are identical featureless spheres. That is the assumption the next unit removes.
Going further
There is a version of this that goes too far. "Scientists are always changing their minds, so you cannot trust any of it" is a conclusion people reach from exactly the evidence on this page, and it does not follow. Look at what actually survived each revision: Dalton was wrong that atoms cannot be split, and every single thing he used that claim to explain is still explained the same way today. Newton was superseded by Einstein, and NASA still uses Newton to land spacecraft. Replacement in science is almost never demolition — it is a new model that has to reproduce everything the old one got right, plus the thing that broke it. That constraint is why the changes accumulate instead of cancelling out.
Before this lesson
Next unit
- Why ice floats
Nearly everything shrinks when it freezes. Water does the opposite, which is why ice floats instead of sinking — and it is the hardest thing this model has to explain.
- The atom: Dalton's model
At GCSE this becomes
- Bonding and structure — where every failure on this page gets its proper explanation.
Where to next
- Next: The atom: Dalton's model
Atoms, elements and compounds
- Previous: Diffusion
Ask Mr Badmus AI
Unsure how a model can be wrong and still be right to use?
Lesson content © MrBadmusAI.